Comparison Of Cycling Crank Power Measured By Instrumented Pedals And A Commercial Power Meter

The measurement of crank power on a bicycle has many applications from providing feedback to cyclists in training to assisting a physical therapist in measuring patient progress. These measurements were restricted to laboratory settings until 1986, when mobile power meters became widely available to measure crank power. These commercial systems generally have a claimed accuracy of plus or minus 2%, but these claims commonly do not have sufficient scientific backing. PURPOSE: To assess measurement discrepancies between crank powers measured by instrumented pedal (gold standard) and a commercial power meter. METHODS: Nine healthy subjects (21 ± 2 yrs, 6 Female) performed three separate two-minute cycling conditions at 80 revolutions per minute at workloads of 1 kg, 2 kg, and 3 kg on two different sets of pedals on a Monark cycle ergometer for a total of six conditions. Crank power from commercial power measuring pedals (CP) was sampled at 1 Hz via bike computer and the pedal reaction forces from force measuring instrumented pedals (IP) were sampled at 1200 Hz. Using derived crank torque, crank angular velocity and pedal angle, crank power was calculated for the IP and compared to power reported by the CP. CP provides average crank power per second and the IP average crank power was obtained per pedal cycle (0°to 360°). Bilateral crank power was computed using the CP provided left-right balance percentage of average total power while the IP are inherently bilateral. Crank power comparisons were made using two-way (pedal x workload) repeated measures ANOVAs. RESULTS: There was a significant effect of workload for total, left, and right crank powers (p < 0.001) and these powers became progressively higher with increased workloads. No significant effects of pedal and interaction were found. Specifically, total IP and CP crank powers at 2 kg (164 W and 166 W) and 3 kg (248 W and 245 W) were higher than 1 kg (96 W and 93 W), respectively (all p < 0.001). The total powers at 3 kg were higher than 2 kg. Similar results were found for left and right crank powers. The highest percent difference between measurements was approximately 6.1% at the right crank power in 3 kg workload. CONCLUSION: No significant differences were found between CP and IP crank power measurements across three workloads, supporting the validity of commercial power measuring pedals.